SUB-NATIONAL CONSTITUTIONS IN FEDERAL SYSTEMS

Michael Burgess and G. Alan Tarr Prepared by the Forum of Federations with the support of the Government of Canada.

SUB-NATIONAL CONSTITUTIONS IN FEDERAL SYSTEMS

Michael Burgess and G. Alan Tarr

CONTENTS

Foreword ...... i 1. Introduction ...... 2 2. Sub-national Constitutionalism ...... 3 2.1 The Sub-national Perspective ...... 4 2.2 Establishing and Maintaining Sub-national Constitutional Space ...... 4 2.3 Supervising Sub-national Constitutional Choices ...... 5 2.4 Preempting Sub-national Constitutional Choices ...... 6 2.5 Reviewing Sub-national Constitutional Choices ...... 6 2.6 Different Scopes of Sub-national Constitutional Space ...... 7 3. Using the Sub-national Constitutional Space ...... 10 3.1 Political factors in sub-national constitutions ...... 10 3.2 Consequences of Using Sub-national Constitutional Space ...... 12 4. Case Studies ...... 14 4.1 The United States, Germany, and Austria ...... 14 4.2 Bosnia and Herzegovina, Switzerland, Belgium, and Canada ...... 15 4.3 The United Kingdom, Spain, and Italy ...... 18 4.4 Supranational integration in Europe ...... 20 Endnotes ...... 23

FOREWORD

The right to have subnational constitutions has been a constant demand of Myanmar’s ethnic minorities. This is perhaps driven by the perception that subnational constitutions will act as an enabler for the more robust protection of human rights and minority rights across the country, and particularly in the ethnic states. This demand therefore is neither new nor frivolous, having been discussed over more than six decades of conflict in the country.

In late 2019, State Counsellor Daw Aung San Suu Kyi, assured the ethnic minorities the right to subnational constitutions in her address commemorating the 4th anniversary of the signing of the Nationwide Ceasefire Agreement. Outlining the government’s determination to work towards a federal union, she mentioned that “States will have a constitution that assures self-determination”. This shows the importance of subnational constitutions as a major pillar in Myanmar’s federalization.

The question of self-determination of ethnic minorities also features prominently in the Federal Democracy Charter promulgated by the Committee Representing Pyidaungsu Hluttaw (CRPH) on April 1, 2021, in response to the February 1 coup by the Myanmar military. What is further required is defining of what would be the framework within which the constituent units in Myanmar would frame their constitution.

The Forum of Federations has been supporting dialogue around federalism in Myanmar for almost a decade now. In the spirit of “learning from each other” we bring to the table international experiences – we do not hold normative views on subnational constitutions or other aspects of federalism.

This paper forms part of the Forum of Federation’s current project “Strengthening Federalism and Inclusive Governance in Myanmar” which is funded by the Government of Canada. The paper draws extensively on the work of Alan Tarr and the late Michael Burgess which culminated in the publication of Constitutional Dynamics in Federal Systems. Sub-national Perspectives. It synthesizes their main research findings, which have been updated by Giacomo Delledone.

The purpose of this paper is not to comment on or advocate a particular model or approach for Myanmar. Rather it presents the state of the art in comparative international practice, which we hope will stimulate debate and discussions on the way forward and options for Myanmar.

Felix Knüpling Vice President, Forum of Federations

i

1. INTRODUCTION

The capacity to respond to change and the challenges associated with it are crucial to the success and survival of federal states. The constitutional arrangements within a federal system for dividing power, resolving disputes, safeguarding rights, and providing for reform and renewal are key in responding to these challenges. This article examines how the constitutional architecture of federal and quasi-federal systems has influenced their evolution and development, their success and survival. This article also looks at the constitutional architecture of these systems from “below”—from the point of view of sub-national constitutions and the regions that they govern.

To many readers, the term “constitutional architecture” may sound odd, but it points to an important overlooked feature of federal constitutionalism. In every federal system, political arrangements at the national level are structured by a federal constitution. In some federal systems, the federal constitution prescribes the political institutions and processes for the country's constituent units—such as provinces, states or cantons—as well, thus dictating the constitutional architecture for the entire federal system.1

But in most federal systems, the federal constitution is acknowledged as an incomplete framework document, in the sense that it does not prescribe all constitutional processes and arrangements. Rather, it leaves space to be filled by the constitutions of its sub-national units. The scope of this sub-national constitutional space varies from one federal system to another. Typically, the less detail that the federal constitution provides, the greater the space for sub-national constitutional space. Usually, federal constitutions also set parameters that constrain the choices available to those drafting sub-national constitutions.2 Nonetheless, this system of dual constitutionalism furnishes an opportunity for constituent units to define their own goals and establish their own governmental institutions and processes. This is a part of the self-rule that is fundamental to a federation.

Sub-national constitutions also form part of the overall constitutional framework for the federation. They are “part and parcel of the total constitutional structure of federal systems and play a vital role in giving the system direction.”3 Constituent units may make use of the constitutional space available to them to initiate reforms to respond to the problems they confront. These experiments, if successful, may have consequences beyond their own borders. A successful experiment in one constituent unit may promote emulation by other constituent units that confront the same problems: a process known as horizontal federalism. In addition, successful experiments in the constituent units may encourage the federal government to adopt the same reforms at the national level, so that the pattern of influence is bottom-up rather than top-down. Thus, to fully understand constitutional evolution and development in federal and quasi-federal systems, one must constantly be mindful of the interplay of federal and sub- national constitutions.

2

2. SUB-NATIONAL CONSTITUTIONALISM

What qualifies a document as a sub-national constitution? In some federal systems—for example, Australia, Germany, and the United States—the question is easily answered. Each federation has distinct identifiable documents that structure political life in its constituent units. In other federal systems, such as Belgium and Nigeria, the question is likewise easily answered, because there are no documents distinct from the federal constitution which perform that function.4 However, in some federal systems the answer is not so clear. For one thing, constitutions in some federal systems are not enshrined in a single document. In Canada, for example, several documents comprise the federal constitution, and the same is true for provincial constitutions.

The situation is further complicated by the emergence of supra-national entities such as the European Union. In some ways, the interactions between the national governments of EU members and the EU resemble those between federal governments and their constituent units. Beyond that, the constituent units of federations find themselves subject to two levels of authority—they can violate neither EU law nor national law in the constitutional arrangements they devise or the policies they pursue. Yet because sub-national units have a consultative role in EU policy through the EU’s Committee of the Regions, the influence is bottom-up, as well as top-down.

Finally, sub-national constitutions might be different in kind from the national constitution. For example, the “Autonomy Statutes” in Spain exhibit attributes of constitutions and structure political life within constituent units but are not called constitutions.

Asymmetrical federal systems pose a particular difficulty. Some federal systems allow only some constituent units to devise sub-national constitutions. For example, the Indian Constitution permits only one state, Kashmir, to have a sub-national constitution. Even that state’s constitution has been rendered inoperative by a Presidential Order in 2019 abolishing that state’s special status and making all provisions of the Constitution of India applicable to it. Other federal systems have special arrangements for some constituent units for historical reasons. In Italy, for example, the province of South Tyrol has long enjoyed a wider autonomy than other regions. This special treatment was based on the international agreement that transferred this territory from Austria to Italy after World War I.5 Other federal systems grant different constitution-making authority to various constituent units. For example, by statute the United States Congress has authorized some territories, such as Guam and the Virgin Islands, to draft their own constitutions. However, unlike the states, these territories are obliged to submit their proposed charters to Congress for approval or revision before they can be submitted to their residents for ratification. American Indian tribes can also devise their own constitutions, but both congressional legislation and their status as “internal dependent nations” circumscribe the constitutional choices available to them.6

How should a country with an asymmetrical federal system determine what qualifies as a sub-national constitution? There are two possible answers to this question.

3

• Make the determining factor whether the federal government or the sub-national unit has the legal authority of enactment, or • Make the determining factor involve the entire range of discretion available within a sub-national unit in designing its governmental processes and institutions.

This study uses option 2 above—a wider scope of sub-national constitutionalism, one reflective of the diversity of current and developing political practices.7

2.1 THE SUB-NATIONAL PERSPECTIVE

Constitutional federalism is usually studied from the vantage point of the federal constitution. Analyzing constitutional arrangements in such a top-down fashion encourages one to frame questions from the perspective of the federal government:

• From the federal perspective, there are three main questions: • What powers does the federal government have? • What constitutional impediments stand in the way of it achieving its objectives? • How far does the federal constitution go to mandate constitutional structures for sub-national units?

Much less common is the view from below.7 Indeed, sub-national constitutions remain low-visibility constitutions to citizens and scholars alike. Yet the sub-national perspective yields insights that the national perspective does not.8 A sub-national perspective also highlights four questions that the national perspective does not:

• What range of discretion—constitutional space—is available to constituent units in designing their constitutional arrangements, and how are the boundaries of that space policed? • What factors influence the scope of sub-national constitutional space in various federal systems? • Why have sub-national units occupied or failed to occupy the constitutional space available to them? • What have been the consequences of sub-national constitutionalism on horizontal and vertical relations within the federal system?9

2.2 ESTABLISHING AND MAINTAINING SUB-NATIONAL CONSTITUTIONAL SPACE

Taking a sub-national perspective requires an assessment of the scope of sub-national constitutional space. This space is the autonomy that the constituent units are allotted within the federal system. This space varies dramatically from one federation to another. If one thinks of constitutional space in terms of a continuum, Switzerland and the United States would be at one end, as systems that permit constituent units wide leeway, and South Africa and Sudan would be at the other end, as systems in which sub- national constitutional space is very restricted.10 The scope of this sub-national space is typically determined by federal law—in particular, by the federal constitution—as well as by federal statutes and

4

administrative regulations insofar as they preempt state law. However, some federal systems provide for a limited sub-national participation in defining sub-national constitutional space. For example, Section 33 of the Canadian Charter of Rights and Freedoms—the so-called “notwithstanding clause”—permits provinces to act despite acknowledging that their action violates a provision of the Charter. To that extent, the provinces determine the range of action available to them.

2.3 SUPERVISING SUB-NATIONAL CONSTITUTIONAL CHOICES

The next question is how federal systems ensure that constituent units respect the outer limits of their sub-national constitution-making space. One way—with minimal constitutional conflict—is for the federal constitution to give the federal government some control over the content of sub-national constitutions at the time they are created. This, of course, requires that the federal government predate the creation of those constitutions, and this is not always the case. In the United States, for example, the thirteen states that declared independence from England in 1776 devised their initial constitutions prior to the adoption of the nation's first constitution. There was no way a federal authority could impose conditions on what would be contained in the original thirteen states’ constitutions. Most of the other thirty-seven states, however, were formed from territory governed by the United States, with Congress controlling the admission of states.11

Implicitly, Congress is authorized by the United States Constitution to establish the conditions under which new states will be admitted.12 Thereby, Congress imposed conditions on what state constitutions should contain in the acts by which it authorized prospective states to devise constitutions and apply for statehood. If a proposed constitution contained provisions of which Congress or the President disapproved, either of them could refuse to admit the state until the offending provisions were removed. This served a deterrent function. State constitution-makers refrained from including in their charters any provisions that might excite opposition in Congress and jeopardize admission to the Union.

The system just described is not common to all federal countries.13 Germany, for example, imposed no special requirements on the constitutions drafted by the five Länder governments that became part of the country following the collapse of the German Democratic Republic.14 However, in countries in which the federal legislature has responsibility for crafting the functional equivalent of the sub-national constitution, such scrutiny is built into the ordinary process of legislation. This is true, for example, in China and Italy. Switzerland requires that the Federal Parliament “guarantee” that the cantonal constitution be consistent with federal law. Article 155 of the Spanish Constitution authorizes the government to take all necessary measures to ensure that self-governing communities, such as Catalonia and the Basque Country, fulfill the obligations imposed on them by the constitution and other laws. This became the basis for the Spanish Government’s suspension of Catalan political autonomy in 2017, when the Catalan government scheduled an illegal referendum on Catalan independence.

5

2.4 PREEMPTING SUB-NATIONAL CONSTITUTIONAL CHOICES

Another way to minimize conflicts between federal and sub-national constitutions is to prescribe the contents of the sub-national constitutions in the federal constitution. Indeed, the federal constitution may obviate the need for sub-national constitutions altogether by mandating the form of government for sub- national units, thus eliminating all sub-national constitutional space. Formerly unitary countries that have decentralized into federal systems are particularly likely to include sub-national constitutional arrangements in the national constitution.15

Even if a country permits the creation of sub-national constitutions, mandates in the federal constitution can restrict the range of choice for sub-national constitution makers or induce the sub- national units to alter their constitutions to bring them into conformity with national requirements.16 For example, the Brazilian Constitution mandates in detail the content of state constitutions. Brazil’s Constitution even specifies the number of state legislators and their pay ceilings. The Mexican Constitution prescribes a separation of powers in state government and limits state governors to a single six-year term.17 A less specific kind of mandate is the homogeneity clause found in the German and Austrian federal constitutions. It requires the constitutional order in the Länder to conform to the principles of a republican, democratic, and social state of law.18 And the Austrian Constitution further requires that Länder constitutions not “affect” the federal constitution.19

In many federations, federal constitutional provisions not directly addressed to sub-national constitutions also have an important effect on state constitution makers. Perhaps the most important of these is the supremacy clause found in most federal constitutions. These supremacy clauses confirm that federal law is superior to state law, so that in cases of conflict, valid federal enactments—be they constitutional provisions, statutes or administrative regulations—prevail over state enactments, including state constitutional provisions. This limits sub-national constitutional space, and it may deter sub-national constitution-makers from adopting some provisions they favor. Just as important are the lists of competences awarded either exclusively or concurrently to the federal government. The broader the range of competences granted exclusively to the federal government, the fewer the opportunities available to sub-national units to address matters themselves. The broader the range of concurrent competencies, the greater the opportunities for the federal government to occupy the entire field.

2.5 REVIEWING SUB-NATIONAL CONSTITUTIONAL CHOICES

Mechanisms for policing or resolving constitutional disputes are just as important as strategies for preventing them. One widely used mechanism for policing the boundaries between the federation and its constituent units is review by federal supreme courts or constitutional courts. In the United States, the Supreme Court has declared several provisions of American state constitutions unconstitutional under the federal Constitution, even though the federal Constitution contains few express restrictions on the states' constitution-making powers.20 In South Africa, the Constitutional Court rejected the entire draft provincial constitutions of Kwazulu-Natal and the Western Cape Province. In the former case, it was noted that some provisions in the proposed constitution appeared to have “been passed by the KZN legislature under a

6

misapprehension that it enjoyed a relationship of co-supremacy with the national Legislature and even the Constitutional Assembly.”21

Although most federal systems rely on the judiciary to resolve competency disputes, this is not the only possible approach. The Constitution of the Russian Federation, for example, authorizes the president of the Federation to suspend the acts of sub-national executives if he believes them to be in violation of federal law or human rights. The Justice Ministry also can revoke regional laws that are in violation of the Federation Constitution.22

Finally, in addition to legal mechanisms for resolving disputes about the scope of sub-national constitutional space, there are a variety of informal mechanisms for resolving disputes about constitutional authority. For example, in Spain, the legal/litigation model of resolving competency disputes between the national government and the autonomous regions was for some time transformed into a process of consultation and negotiation, until the 2017 conflict over Catalan independence.23 In Italy, by contrast, which is at a much earlier stage of creating and recognizing regional autonomy, the legal/litigation model remains dominant. These examples suggest the transition from the legal/litigation model of policing sub-national constitutional space to the consultation/negotiation model may be a hallmark of a maturing federal system.24

2.6 DIFFERENT SCOPES OF SUB-NATIONAL CONSTITUTIONAL SPACE

If the scope of sub-national constitutional space varies among federal systems—and sometimes even within federal systems—what produces this variation? Several factors may be relevant. The process by which the federal system was created would be one such factor. Scholars distinguish between federal systems created by uniting pre-existing political entities, known as aggregative or coming-together federations, and those created by the transformation of a previously unitary political system, known as devolutionary or holding-together federations.25 This dichotomy oversimplifies somewhat, as there are federations whose formation has involved both aggregative and devolutionary processes. For example, the Swiss Federation was formed by the merger of preexisting political societies, but the subsequent creation of the canton of Jura could be seen as devolutionary. Still, the distinction remains useful.

One would expect aggregative federal systems to allow more sub-national constitutional space than would devolutionary federal systems. In part, this would simply be the product of historical context. When political units form a federation, they already have in place their own institutions and political practices. Attempts to interfere with them or to prescribe unnecessary uniformities might threaten the process of federation. In addition, the political units joining together to create a federation are likely to seek to retain self-rule to the extent consistent with achieving the ends of federation. Prospective constituent units might also demand concessions expanding or safeguarding sub-national constitutional space as the price for joining the federation. For example, several southern states threatened not to join the American Union unless states were free to determine their own laws on slavery. Finally, as a matter of constitutional design, aggregative federal systems are more likely to lodge residual powers in the constituent units rather than the federal government.

7

A different dynamic is likely in devolutionary federations. The national authority would be unlikely to surrender powers beyond those necessary to achieve the ends of federation. Moreover, constituent units that are being created rather than pre-existing do not have the ability to make demands as to the scope of sub-national constitutional space. Often, they lack a strong political identity—sometimes intentionally so. When South Africa created its nine provinces, for example, it split the provinces that had constituted the original Union of South Africa and incorporated the homelands established by the apartheid government. In that way, it drew provincial boundaries in a way that most provinces were ethnically heterogeneous, which dissipated the power of ethnically based political groups.26 Finally, in contrast to aggregative federal systems, devolutionary systems are likely to lodge residual powers in the federal government, thereby circumscribing the powers—including constitution-making powers—of the constituent units.

Another factor that might influence the scope of sub-national constitutional space is whether the federation has a system of symmetrical or asymmetrical federalism. In a symmetrical federal system, all constituent units have the same powers of self-government, but in an asymmetrical federal system one or more constituent units are vested with special or greater self-governing powers. Federations usually create asymmetrical arrangements to “take account of the fact that within a state there are significant cultural or societal differences among the constituent units.”27 This is particularly important when there are groups that desire a degree of autonomy but are destined to be permanently in the minority at the national level. Incorporating asymmetrical elements may reduce the conflict that this could produce by allowing minorities concentrated in particular constituent units a greater measure of self-rule, thereby wedding them more closely to the federation.

This greater self-rule and the recognition of the diversity that led to the asymmetrical arrangement in the first place would usually require extensive sub-national constitutional space. Yet it may be difficult to limit such self-rule to the distinctive constituent units within the federation. Other constituent units might well resent the “privileges” that are given—think, for instance, of the reaction of the “Rest of Canada” (the other nine provinces plus three territories) to the claims of Quebec—and demand the same opportunity for self-rule.28 But whatever the eventual outcome, one would expect that there would be broader sub-national constitutional space in asymmetrical federations.

A further factor affecting the scope of sub-national constitutional space might be the purposes underlying federation. Some federations—such as Switzerland, Nigeria, and Belgium, as well as quasi- federations such as Spain—were designed to recognize and accommodate the multi-ethnic character of the population and provide space for the expression of diversities.

One would expect in such instances that the constituent units would largely correspond with the diversities within the population and that the federation would accord broad constitutional space to the constituent units. This expectation is only partially borne out: although constituent units do mirror the political saliency of ethnicity in the federations, neither Belgium nor Nigeria has sub-national constitutions, and Catalonia and the Basque Country in Spain have only autonomy statutes. Also, some federations, such as Switzerland, provide broad sub-national constitutional space, but others, such as

8

Malaysia, do not. In those that do not, often greater representation in the councils of the federal government substitutes for self-rule.

Finally, some federations or quasi-federations have been designed to deemphasize the ethnic or religious divisions in the society. The goal was to replace fragmentation with national solidarity and a common national identity. In such federations broad sub-national constitutional space may be seen as a threat to national unity. This is particularly a concern if (as in India) the boundaries of current constituent units reflect the language groupings within the population. Thus, it is hardly surprising that in India and South Africa, two prime examples of multi-ethnic federations committed to forging a common national identity, there is little sub-national constitutional space. Also, the national governments in India and South Africa are authorized to invade even those powers that the federal constitution gives exclusively to the constituent units when this is necessary to serve the purposes of national economic unity, national security, and the need for national uniformity.29

Most federations are not focused primarily on dealing with ethnic or religious diversity. Some countries, such as Argentina, Brazil, and the United States, have embraced federalism primarily as a way to govern large territorial expanses. In such circumstances, one might expect that constituent units would be granted broad constitutional space in order to permit locally appropriate responses to diverse conditions. Other countries, such as Austria and Germany, have embraced federalism as a way to promote administrative efficiency, with the constituent units primarily responsible for implementing federal policy.30 One would expect that such federations would emphasize concurrent rather than exclusive powers and accord their constituent units very limited constitutional space.

These expectations are only partially fulfilled. Whereas the American states do have broad constitutional space, the same is not true for constituent units in Argentina and Brazil. In Brazil, the detailed 1988 Constitution and judicial rulings have virtually eliminated state experimentation in constitution-making.31 On the other hand, although the homogeneity clauses in the German and Austrian constitutions and their emphasis on concurrent powers have limited constitutional experimentation in the Länder, they have not foreclosed it. The German Länder, for example, have adopted constitutional provisions for referenda and since the late 1980’s they have also revised their constitutions, following the lead of Schleswig Holstein, to identify goals for state activity and to expand protections for social rights.32 This suggests that cooperative federalism does not necessarily preclude significant use of sub-national constitutional space.

9

3. USING THE SUB-NATIONAL CONSTITUTIONAL SPACE

Law defines the formal constraints on sub-national constitutional space, but to what extent do constituent units occupy—or fail to occupy—the constitutional space allotted to them?

Four points should be made at the outset.

First, determining whether constituent units have made use of the constitutional space available to them is somewhat tricky. To do so, one might look for differences between sub-national constitutions and the federal constitution, as well as for differences among sub-national constitutions within a federal system. These differences would indicate that the constituent units had in fact considered alternative constitutional arrangements rather than thoughtlessly copying what had been adopted elsewhere. However, this is not foolproof. Constituent units may seriously consider alternatives to what is found in the constitutions of other constituent units or in the federal constitution but conclude that there is no reason to diverge from those models. This would qualify as occupying the constitutional space, because the constitution-makers had made a self-conscious choice rather than merely copying what they found.

Second, occupying constitutional space is not an either/or proposition: constituent units may make use of some, but not all, of the space available to them.

Third, constituent units within the same federation may vary in the use they make of the constitutional space available to them, and this variation may occur in both symmetrical and asymmetrical federal systems.

Fourth, political factors ultimately determine the use of sub-national constitutional space. These factors range from the prevailing political ideas of the era to the nature of the party system, to the level of dominance of a particular party throughout the country, to the nature of popular demands upon sub- national governments.

3.1 POLITICAL FACTORS IN SUB-NATIONAL CONSTITUTIONS

The similarities and differences among sub-national constitutions, as well as their similarity to or divergence from the federal constitution, may reflect the political era in which they were written. Since different sets of political ideas tend to be dominant at various points in time, sub-national constitutions are likely to reflect the reigning ideas of the era in which they were written.33 One sees this in the constitution-making in the German Länder. The Länder constitutions that preceded the adoption of the German Basic Law tended to include “the whole array of political and social provisions, including basic human rights.”34 Those drafted after the adoption of the Basic Law focused on organizational principles, because social concerns and rights guarantees had already been dealt with in the Basic Law. Finally, the Länder constitutions drafted since 1990 have sought to guide political practice through the inclusion of social rights and state goals.35

10

Yet changing political ideas are likely to lead constituent units to make use of the constitutional space available to them only if there is some time lag between the adoption of the federal constitution and the adoption of its sub-national counterparts. This time lag is necessary to allow for a shift in political ideas to take place. In many federations, this is simply not the case, either because the federation is of relatively recent origin (e.g., Russia and South Africa) or because the federation has adopted a new constitution in the recent past (e.g., Argentina, Brazil, Nigeria, and Switzerland).

Also, insofar as the federal constitution can be relatively easily changed, the federation may itself respond to changing political ideas with constitutional amendments. Such amendments keep the federal charter up to date and reduce the need for sub-national constitutions to take the lead in pioneering new directions. On the other hand, if the federal constitution is substantially more difficult to change than the sub-national constitution, over time their contents are likely to diverge even if the two constitutions were adopted at the same time. This may also have broader implications, as the frequency or infrequency of constitutional change may affect how political actors view the constitutions that are amended or revised. In the United States, for example, the infrequency of formal constitutional change at the national level has imbued the federal Constitution with a sense of untouchability, whereas the frequency of amendment as the state level has encouraged the public to view changes in state constitutions as merely part of normal politics.36

Another factor encouraging constituent units to occupy the constitutional space available to them may be regional differences reflecting distinctive political or legal cultures or traditions, sometimes linked to ethnic diversity.37 The diversity in many federations can lead constituent units either to enshrine their residents’ distinctive culture in their constitutions or to provide additional protections to ethnic minorities situated within their borders.

Sometimes one change at the sub-national level may precipitate other changes as well. This is particularly likely when the establishment of new avenues for sub-national constitutional change empowers groups that had previously been stymied. A prime example is adding direct democracy to sub- national constitutions. The availability of this new avenue of change may enable groups that had been relatively ineffective to pursue their objectives, thereby opening the possibility of a succession of constitutional amendments. More generally, the more numerous the mechanisms for instituting constitutional change, the more likely such change will happen. Thus, it will be more likely that constituent units will occupy the constitutional space available to them.

Finally, the distribution of political forces within the federation affects the likelihood that constituent units will occupy the political space available to them. This happens by creating either incentives or disincentives for political mobilization for sub-national constitutional change. If the party that is in control at the national level is in control within the various constituent units, then it is more likely that constitutional reform will be pursued at the national level or that constituent units will model their constitutions on the federal charter.

While the Partido Revolucionario Institucional held power both nationally and within the Mexican states, centralization of power was the norm, and federalism and sub-national constitutions were largely

11

ignored. Conversely, if political parties that are in political opposition at the national level control the governments of some constituent units, they will likely make use of that political control to advance their own agenda, and this may include constitutional innovations in the space available to them. Thus, when the Progressives gained control of the California government in the early part of the twentieth century, they constitutionalized some reforms that were anathema to the conservatives who dominated the federal government.38

The existence of strong national political parties may also discourage distinctive initiatives from constituent units, reducing their interest in occupying the constitutional space available to them. Indeed, some constituent units may make deliberate, rational choices not to occupy fully the space legally allotted to them. South Africa provides an example of this. The African National Congress, as a matter of party policy, mandated that the provincial governments it controlled should not draft provincial constitutions, with the result that only Western Cape Province now has a provincial constitution.39 Conversely, the existence of regional or ethnically based parties might have the opposite effect.39

3.2 CONSEQUENCES OF USING SUB-NATIONAL CONSTITUTIONAL SPACE

Our inquiry into the consequences of constituent units occupying their sub-national constitutional space can be understood as a sub-category in the rich literature documenting the diffusion of innovations inside federations.40 This goes back to Louis Brandeis’ depiction of states as the laboratories of democracy. This depiction assumes that a multiplicity of policy experiments would be more likely to discover good public policy than would a single effort. If the experiment in one constituent unit failed, the damage would be limited to a single jurisdiction. But if it succeeded, then other jurisdictions could emulate the successful experiment in their own law and public policy.

Sub-national constitutional innovations have effects both horizontally and vertically. Existing sub- national constitutions serve as models, either positive or negative, for constitution-makers in other constituent units. This is hardly surprising. The practice of drawing upon or copying provisions reflects in part a respect for the efforts of earlier constitution-makers. Describing the evolution of American state constitutions, Willard Hurst noted41 the willingness to draw upon the experience of other states is enhanced by the recognition that constituent units face common constitutional and policy problems. In symmetrical federal systems, the constituent units share the same powers and confront the same policy concerns, so they tend to be open to what has worked in other constituent units.

Indeed, there is even evidence that borrowing of constitutional innovations may on occasion extend beyond the borders of a single federation. For example, the initiative and referendum provisions added to the Oregon Constitution in 1902 were based on similar provisions found in the constitutions of the cantons of Switzerland.42

Yet when conditions and values differ within a federation, it is less likely that constituent units will emulate the sub-national constitutional innovations pioneered in different units. Thus, when constituent units are organized to reflect differences within the population of a federation, those differences—and the attempt to give them constitutional expression—may lead to the creation of distinctive constitutional

12

provisions that are only appropriate within the particular unit. And if some constituent units in an asymmetrical federal system have greater constitutional space than others do, that will also retard the diffusion of constitutional innovations.

When constituent units occupy the constitutional space available to them, this may also have effects on constitutional politics at the federal level, because the process of imitation and emulation can work vertically as well as horizontally. This happened, for instance, when the United States’ Constitution was drafted. The framers borrowed from some state constitutional provisions but also rejected others. States occupying constitutional space with successful innovations may encourage the federal government to adopt those innovations. But when the federal government does so, federal law may supersede state law and thereby diminish the scope of sub-national constitutional control.

The American experience also reveals that when states occupy the constitutional space available to them, this can produce active avoidance rather than emulation. This has occurred when states have sought to occupy constitutional space by creating state constitutional rights broader than what was available under the federal Constitution. One highly publicized example involved rulings by state supreme courts in Massachusetts, California, and Connecticut recognizing same-sex marriage as mandated by their state constitutions. Instead of emulation, these rulings prompted actions by other states to prevent the diffusion of these innovations, to preempt similar rulings within their own borders by constitutionally prescribing that marriage is limited to male-female couples. The rulings also prompted an unsuccessful effort to define marriage in the federal Constitution, an attempt to federalize the issue not to follow the states’ lead, but to circumscribe state constitutional space. Eventually the issue was resolved when the U.S. Supreme Court recognized the right of same-sex couples to marry under the federal Constitution, but state courts provided the leadership that prompted that decision.43

13

4. CASE STUDIES

The following analysis considers a wide variety of constitutional sub-national experience from a number of comparative perspectives. These short case studies show that constitutional sub-national autonomy in federal states and federal or quasi-federal political systems has contributed to their change and development. The studies explain how and why they have managed to evolve and adapt in both similar and different directions. In the last decade, the impact of the economic and financial crisis that began in 2008 and recurring secession crises in some jurisdictions have clearly affected the discussion about federalism and constitutional sub-national autonomy. Likewise, the unfolding of the Covid-19 crisis is likely to affect the balance of power in federal and quasi-federal jurisdictions—however, it is too early to make any predictions about this.

4.1 THE UNITED STATES, GERMANY, AND AUSTRIA

First, sub-national constitutionalism in territorially based federations is considered. These include the United States of America (USA), the Federal Republic of Germany and the Republic of Austria. It is interesting to look at the USA with the concept of “constitutional space” in mind—the constituent state governments have the opportunity to operate autonomously largely because of the federal constitution’s brevity and its narrow focus upon the structure of the federal institutions and enumerated powers. Their combined impact has left unoccupied a potentially large and expansive area for state government activity. State constitutional processes, for instance, have been an important avenue for promoting political change in the design of governing institutions, in the regulation of the suffrage, in the extension of individual rights and in public policy reforms.

In Germany, the political pressures for more “constitutional space” in pursuit of sub-national autonomy in an established federation have a simultaneous top-down and bottom-up character. “Dual federalism”, the division of legislative powers and the administration of federal laws together enabled the federal government to expand its powers and competences in the legislative and administrative arenas at the direct expense of the Länder. This explains the growing contemporary pressures for constitutional reform in terms of a widespread critique of the current federal political system—such pressures peaked in the late 1990s as a result of reunification fatigue. Between the late 1999s and the late 2000s, the drive for a redistribution of competences between the 16 Land executives derived from dissatisfaction with the perceived excessive unitarism of German federalism, the Länder being left with limited financial responsibility and little room for maneuver through the exercise of their legislative competences. The Land parliaments played a significant role in this process. Since the Land executives were already represented in the Bundesrat, where they participated in the making of federal legislation, it was the Land parliaments and local governments that were effectively marginalized and had only a very weak relationship to federal policymaking. It was this sense of political exclusion which prompted a renewed critique of “executive-administrative” federalism in Germany.

14

Efforts to re-examine the operation of the federal system brought into sharp focus the interrelated questions of territorial reform, legislative competences, the role of the Bundesrat and major budgetary issues regarding federal financial relations. Reform proposals in 2006 and 2009 revealed the conflict between maintaining a commitment to solidarity in a social welfare state while promoting stronger fiscal discipline. This, in turn, exposed the tensions between the provision of more constitutional space for the Land governments and parliaments, coupled with the desire to strengthen “competitive” federalism, and the underlying political economy of German federalism which confirms the economic weakness of the poorer Länder.

In Austria, strong historical and constitutional pressures toward homogeneity among the 9 Länder constitutions, whose commonalities are supposed to harmonize with the Federal Constitution, can be identified. The flip side of this is that the Länder constitutions and governments cannot deviate from the federal legal framework. At first glance, Austria is therefore a highly centralized federation with conspicuous unitary features and this vice-like constitutional grip that the federal power has over the constituent Länder units is one of its obvious hallmarks.

Yet, given a closer look, it becomes clear that the political role of the Länder is stronger than their limited constitutional powers might initially suggest. Characteristic is the pivotal role of the Landeshauptleute (governors of the Länder) as formidable veto players holding key positions in the federal political system. Moreover, since the early 1980s there has been a significant shift in the perception of the Länder constitutions. They are no longer construed as passively subservient but rather as the basic law within the various Länder. In particular, there has been a real recognition that the Länder constitutions can be vehicles of policy and institutional innovation and experimentation, which has had an impact both horizontally among the Länder and vertically on the federal government itself. A few examples deserve mention: a provision concerning the protection of the Slovene minority was entrenched in the Constitution of Carinthia in 2017, while climate protection was included into the constitutional documents of Carinthia, Lower Austria, Tyrol, and Vorarlberg. Indeed, this contemporary trend creates a political opportunity for the Länder to strengthen their “relative constitutional autonomy” within the general framework that is applicable to both the Federation and the Länder.

4.2 BOSNIA AND HERZEGOVINA, SWITZERLAND, BELGIUM, AND CANADA

Second, we look at federations that are primarily characterized by their social heterogeneity. Bosnia and Herzegovina (BiH) is a state built upon a series of paradoxes. In BiH, the drive for change and development has not been brought about internally by sub-national constitutional means used by constituent state units or federated entities but has been imposed by external forces and international actors in the wake of the Bosnian War. Consequently, a remarkable evolution took place—a largely ethnically-based and territorially decentralized diarchy turned into a putative multinational federal political system in which the weak central (federal) authority remains at the mercy of its two constituent units, the Federation of Bosnia and Herzegovina and the Republika Srpska.

15

The idea of constitutional space is therefore turned on its head. Together, the institutionalization of ethnicity and the partition of power along territorial lines have solidified a bottom-heavy ethnic federalism which defiantly refuses to give up its powers and competences to a fragile central authority. For sure, the situation present in BiH challenges conventional ideas of constitutional and institutional engineering and design. In BiH, three distinct historical processes appear to be evident: competitive state building at two levels, nation building in terms of three ethnic nations at the level of the discrete entities, and the larger multination building at the federal level. From a comparative perspective, BiH comes close to the case of the European Union (EU) with strong constituent units in the member states and a relatively weak center in Brussels. However, it stands in stark contrast to the case of Belgium, often associated with a hollowing out of the center by the two main linguistic constituent units, where contemporary trends suggest a movement in the opposite direction.

Switzerland is a good example of constituent sub-national autonomy rooted in the 26 cantonal constitutions. This multilingual, multicultural federation is still revered as a model of federal order and stability in conditions which would appear to be unpromising. The constitutional world of the cantons considered from a sub-national perspective connects directly with the renowned social diversity of the federation.

A range of reasons and arguments can be identified to have triggered recent constitutional revisions and amendments in each of the cantons. Rather like the cases of the US states and the Austrian Länder, the Swiss cantons have historically taken constitutional and political initiatives well in advance of the federal authorities, especially in terms of popular rights and liberties and in their role as laboratories for democracy and good governance. In so doing, the cantonal constitutions help fulfill the important role of self-rule in the federation, meet the contemporary challenges of political life and reinforce the legitimacy of cantonal law and authority in the daily life of Swiss citizens. This links them closely to the socio-political, legal and ultimately federal reality. For the sake of completeness, however, it should be added that cantonal constitutions sometimes proved to be less receptive to social change. Some cantons, most notably Appenzell Innerrhoden, were remarkably slower than the federal government in introducing female suffrage.

Next, the question of the constitutional and institutional autonomy of the regions and communities in Federal Belgium is addressed. This country is based upon three main linguistic communities, namely the Dutch speakers (Flanders and a tiny minority in Brussels), the French speakers (Wallonia and Brussels) and the German speakers (Eupen and Malmedy in Wallonia) which are largely territorially concentrated. Belgium is a typical example of devolutionary federalism. The implications of its gradual movement from a long established unitary but territorially decentralized state to a new federation have influenced the relationship between the federal authority and the various federated entities (regions and communities). From a formal legal perspective, the constituent regional and community legislature have never had the sort of constitutional sub-national autonomy that we can find in most federations, such as in Germany and the USA (integrated federalism). This is largely due to the absence of any residual powers allocated to them. This power has been retained by the federal government, while the constituent units have only the enumerated powers awarded to them in the federal Constitution.

16

Therefore, the federated constituent units enjoy only a form of “constitutive autonomy” rather than any substantive sub-national constitutional autonomy; this certainly does not furnish a robust legal basis sufficiently developed for the regions and communities to adopt their own constitutions. Their status is already firmly entrenched in the federal Constitution, which serves as the legal basis for the federal state, the federal government, and the regions and communities. In a nutshell, “constitutive autonomy” allows the regions and communities to introduce deviations from the federal legislative framework. Due to the deep asymmetries that characterize Belgian federalism, the scope of the “constitutive autonomy” used to vary from one sub-national unit to another. However, the latest major reform of the Belgian Constitution in 2013 somehow reduced the gap and extended the scope of the “constitutive autonomy” of the German Community and the Brussels Region. Although this is the exact opposite of a so-called integrative federal model, it has effectively served to accommodate the main linguistic identities that continue to shape the federation. Thus, in Belgium, the process of federalization is a continuous one—as the subsequent waves of constitutional reform since the late 1960s show, the nominal transformation of Belgium from a unitary to a federal state in 1993 was just one stage in this process.

Further insight into the topic of sub-national constitutionalism in multinational federations is gained from looking at Canada. As in Belgium, the constituent units of the Canadian federation do not have formally written sub-national constitutions and much of their constitutional identity is subsumed within the Canada Act (1867). The Canada Act remains the bedrock of the parliamentary federal tradition and serves as the legal basis for the federation, the federal government and the provincial governments. The formation—or reformation—of Canada as a federation in 1867 contained both aggregative and devolutionary elements. This is evidenced in the fact that the residual powers were retained not by the provincial governments but by the central (federal) power. In this way, the Canada Act reinforced the unitary character of the new federal model, combining the Westminster model of parliamentary government with the federal principle.

The peculiar nature of Canada’s federal constitution that combined strong unitary elements with significant decentralist federal features—subsequently given greater, if unanticipated, legal expression— established in practice an ambiguous relationship between sub-national provincial autonomy and federal authority. The lack of formal constitutional entrenchment of provincial political authority has fostered the evolution of a distinctive provincial constitutionalism in Canada. In the absence of formal sub-national constitutional autonomy, the provinces have utilized an opportunist political strategy to promote their own discrete interests. In this respect, negotiations between the provincial executives and the Canadian government have generally played a key role.

The sense of sub-national constitutional space in Canada, then, has emerged gradually through the interaction of largely unwritten provincial constitutions and constitutionalism, and the constitutional culture at the level of the federation. But Canada’s constitutional tradition remains centralized, so that the federal constitutional culture sits alongside an array of distinct sub-national constitutional cultures that are conservative. There has been a marked constitutional reticence by the provinces to define and express themselves constitutionally, except for the provinces with a distinctive cultural or political identity, i.e., Quebec and, more recently, Alberta. The example of Canada encourages us to look beyond

17

the familiar understandings of what is a constitution and at the mixed constitutional heritage by which it is characterized, and which has produced overall governmental and political stability.

4.3 THE UNITED KINGDOM, SPAIN, AND ITALY

Third, our focus next shifts to three territorially decentralized states: the United Kingdom (UK), Spain and Italy. Each of these decentralized states has utilized a series of legal and political practices, that in creating constitutional space and promoting constitutional sub-national autonomy, have developed strong federal or quasi-federal elements. Together, they prompt us, as in the Canadian case, to distinguish not only between constitutional theory and practice but also between constitutional and political practice.

The process of constitutional change in the UK can be described, at least in its first stage, as a “Quiet Devolution”. The marked shift in the UK constitutional culture since 1997 has its roots in a century-old history and in the consolidation of these recent institutional changes. These developments have created opportunities for Wales, Northern Ireland and Scotland to occupy constitutional space and to infiltrate the central organs of the state in order to influence and shape a new constitutionalism, in ways unimaginable only a decade ago.

From a historical perspective, the peculiarly bilateral approach to incremental union building and the different experiences of administrative devolution seem to be the main reasons for the notable asymmetry of the 1998 devolution project. However, the important shift in what could be referred to as cultural autonomy of the territorial societies of Scotland, Wales and Northern Ireland also served to prepare the ground for the relatively smooth political transition to devolution. There was, in this sense, an organic dimension, springing from below, which indicated that a significant change in the constitutional culture was already in place, making it receptive to the emergence of devolution.

The past decade first seemed to confirm the virtues of the pragmatic approach underlying the devolution process. The rejection of the secession option at the Scottish referendum in 2014 seemed to pave the way for “Devo Max”, as the grant of full fiscal autonomy to Scotland was known. Since 2016, the handling of the Brexit issue has put the territorial arrangements that are part of the British unwritten constitution under serious strain. The Remain option prevailed in Scotland and in nationalist-dominated areas of Northern Ireland. Consequently, claims for differentiated Brexit, including the preservation of a “soft border” between Northern Ireland and the Republic of Ireland, have pretty much complicated the task of the British and European negotiators. Meanwhile, stronger pressures for independence from the United Kingdom started emerging in Scotland, Northern Ireland and even in Wales, where nationalist movements used to be much weaker.

Some of the trends and circumstances found in the UK are likewise found in Spain. Looking at the Spanish Constitution and the statutes of autonomy related to the 17 Autonomous Communities (ACs) that together comprise the Spanish state, a series of what can be called “bottom-up structural changes” to the constitutional rules that regulate the organization of territorial power have occurred.

18

Spain has been facing difficulties in undertaking the long overdue reform of the Spanish Constitution. The obstacles to reform include both procedural barriers and political party differences at both national and AC levels. Yet, unlike constitutional reform, territorial reform has not been prevented. However, alternative means exist to adapt the constitutional system, such as legislative and de facto approaches, which have consequently assumed a special significance. In this regard, the statutes of autonomy have played a pivotal role. And thereby, the ACs have taken the lead in the quest to consolidate and adapt the Spanish system of political decentralization.

What we see in Spain is a parliamentary monarchy with an unfinished constitution that needed to be completed principally by its constituent units using their own statutes of autonomy. These were designed to “flesh out” the territorial structure of the larger state. Consequently, the Spanish model has left open to debate a seemingly endless number of legal, political and constitutional questions. The ambiguous and sometimes confusing relationship between what the Spanish Constitution states but does not define, and how far the ACs can legally stipulate their own powers gives the impression that the legal door is still wide open to the expansion of powers and competences of the ACs.

Ultimately, the original constituent power lies with the Spanish Parliament, which with a Spanish- wide purview has the constitutional power to give its initial approval to the statutes of autonomy, and then formally to approve their subsequent reform. The overall position of the ACs, however, is potentially strong in terms of having the strategic capacity to initiate the process of establishing and reforming their statutes of autonomy. This gives them an important influence on the central institutions’ final decision, albeit this is the only way that they are able to participate in the reform of the Spanish Constitution itself.

To understand the process and substance of constitutional sub-national autonomy in Spain, it is crucial to consider the asymmetrical nature of the territorial structure of power relationships with the current political pressures for financial reform. Overall, the leveling out of asymmetry as the ACs gradually acquire similar powers and competences suggests a rather holistic vision of Spain.

This mechanism, based on decentralized constitutional adaptation and moderate asymmetry, has suffered a crisis in the last decade. First came the last wave of reform of the statutes of the ACs, most notably the new statute of Catalonia. When the conservative People’s Party challenged the new statute of Catalonia, the Constitutional Court issued a judgment striking down some of its provisions and deprived many others of any binding legal force. Meanwhile, the economic and financial crisis severely damaged the overall internal balance of the Spanish autonomic state. These two factors coincide in explaining the crisis of the Spanish autonomic state, explicit protest the parliamentary monarchy in nationalist and far- left circles, and the rise of the Catalan pro-independence movement. After a highly problematic plebiscite on independence, the Spanish government went so far as to remove the President of Catalonia and to dissolve the regional legislature. After the snap election, however, the pro-independence political parties have retained much of their political strength in the legislature in Barcelona.

Italy is another example of a formally non-federal but territorially decentralized state. Italy can be called quasi-federal because it has introduced federal elements in the state structure, but there is still no

19

real understanding among mass publics about what being federal means. So, a federal political culture is not completely absent, but fragmented territorially.

History plays a key role in understanding recent developments in Italy. After Napoleonic centralization and fascist totalitarianism, Italy adopted a republican Constitution in 1948 and only implemented its regional prescription in the 1970s. The idea of the regional state is therefore of very recent origin and has, from its inception, been characterized by an asymmetrical design. The creation of 15 “ordinary” regions and 5 “special” regions is often considered as a “third way”, albeit not an easy one, between a conventional federal and a unitary political system. Interestingly, distinctive political cultures, with strong regional parties and emphasis on regional autonomy, are more typical in some of the special regions, most notably those in which historic minorities are present. Such is the case with Trentino-Alto Adige/South Tyrol and the Aosta Valley.

Important constitutional amendments in 1999 and 2001 led to the direct election of the regional presidents, the enhancing of the constitutional sub-national autonomy of the regions and the reorganization of the relations between the national government and the regions. After the reform came into force, the Constitution also opened up to asymmetrical regional autonomy, giving ordinary regions the chance to catch up in some policy areas with the special regions. Hence, a potential leveling out of asymmetry, as in Spain, looked possible.

Unfortunately, two decades after its adoption, the 2001 reform has still not been completed due to financial problems and governmental instability, as well as conflict over some of the necessary implementing reforms. These factors jointly encouraged the Constitutional Court to play an enhanced role in clarifying and determining the precise meaning of the disputed elements. Sometimes judgments have been made in favor of the national government, at the expense of the regions. In effect, the Court has begun to rewrite the division of legislative and administrative powers originally laid down in the reform. In the first decade after the adoption of the reform, judicial interpretation rather than the democratic political process seemed to shape the constitutional future of Italy. Subsequently, the economic and financial crisis and some political scandals triggered a discussion about the actual virtues of having strong regions and, possibly, about a recentralization of powers and resources.

In 2017, the regional governments of Lombardy and Veneto organized consultative referendums in order to launch a process for the grant of greater autonomy under Article 116.3 of the Constitution. Soon after, they were joined by the regional government of Emilia-Romagna. However, political instability in Rome and the subsequent outbreak of the Covid-19 crisis have halted the discussion about asymmetric regionalism.

4.4 SUPRANATIONAL INTEGRATION IN EUROPE

Lastly, the focus shifts from territorially decentralized states to supra-national constitutionalism and constitutional futures in Europe. Thereby the concepts of constitutional space and sub-national autonomy are reconfigured as the sovereign states themselves become the constituent parts of a larger union of states. An essentially two-way process of adjustment and adaptation is brought about by EU membership,

20

i.e., in relation to the transfer of powers that membership entails. Many member states are ill-equipped to address the ramifications of EU membership. Hence, the overall adaptation and adjustment is uneven, patchy and has fostered endless debate about the erosion of national sovereignty in the legal and political relationship between the constituent parts and the larger whole that is the EU. To some extent, the question of national constitutions and the EU has triggered a new debate about the nature of constitutionalism itself—an issue already introduced in a different context in the UK. Thus, the national constitutions have played a role in driving constitutional reforms at the EU level and in consequence, have shaped an emerging constitutionalism in the EU. In dynamic terms, the guiding role of the national constitutions was complemented by the case law of some constitutional courts in the member states. The most influential among them is the Federal Constitutional Court of Germany, which has voiced serious concerns about the actual possibility of conceiving a representative democracy and constitutionalism in a legal environment other than the nation state. A number of major crisis in the last decade, including the sovereign debt crisis, the refugee crisis, Brexit, the apparent decline of constitutional democracy in Hungary and Poland, and the Covid-19 outbreak have all contributed to shaking the balance of power within the European Union.

It is worth examining the internal territorial and constitutional character and choices of the member states, as they adapt their domestic institutions and intergovernmental relationships as a result of EU membership. Looking particularly at the Italian experience, three distinct issues related to the impact of EU membership have arisen: the scope of the constitutional space available to the sub-national units within the state, the extent to which the Italian regions have occupied the constitutional space and the effects of the initiatives of sub-national units on internal changes in the state or to the EU. The concept of “Europeanization” thus fastens on to both top-down and bottom-up perspectives of the constitutional dynamics of change in domestic state structures, that also go hand-in-hand with the process of deepening European integration. Top-down EU pressures have dramatically influenced sub-national member state actors, the competences of the regions, the overall domestic balance of power and, in the Italian case, the pivotal role of the Constitutional Court. The impact also extends to regional implementation of EU law and has important implications for the unity of the state. Regarding the bottom-up perspective of “Europeanization”, several factors assist this process in terms of the representation and participation of both national and sub-national actors in the supra-national decision-making process. This raises inter alia important questions about the constitutional status of the sub-national level and its relationship to national government, the potential enhancement of regional identity and the various mechanisms of access to EU policymaking. EU policies have generally served to strengthen national executives at the expense of national and regional parliaments, thereby altering the constitutional dynamics of sub-national autonomy in Italy.

The EU as the current institutional expression of European integration has, in one way or another, provoked significant constitutional changes in its member states largely, but not solely, in favor of the sub-national units. There are huge changes between and within member states, but their overall impact has been to establish some sort of role for these units in EU affairs through both new and old channels.

21

The general theme of change and development in federations and federal political systems is emphasized by the case studies above, which call attention to the sub-national perspectives of constitutional change and the significance of constitutional space for the evolution of these states and systems.

22

ENDNOTES

1 The use of the term “subnational” is intended to distinguish the constitutions of component units in federal systems from the constitution of the nation state. The authors recognize that many federal systems contain various nationalities, sometimes even called nations, within them.

2 Donald S. Lutz, “The United States Constitution as Incomplete Text,” Annals American Academy of Political and Social Sciences 496 (March 1988): 23, 26; Donald S. Lutz, “From Covenant to Constitution in America Political Thought,” Publius: The Journal of Federalism 10 (Fall 1980): 101-02.

3 Daniel J. Elazar, Exploring Federalism (Tuscaloosa: University of Alabama Press, 1987), p. 174.

4 See Ronald L.Watts, “Provinces, States, Lander, and Cantons: International Variety among Subnational Constitutions,” Rutgers Law Journal 31 (2000): 941-959 and Wouter Pas, “A Dynamic Federalism Built on Static Principles,” in G. Alan Tarr, Robert F. Williams, and Josef Marko, eds., Federalism, Sub-national Constitutions, and Minority Rights (Westport, CT: Praeger, 2004). Of the federations discussed in this study, article, the following have separate sub-national constitutions in most or all of their constituent units: Argentina, Australia, Austria, Brazil, Ethiopia, Germany, Malaysia, Mexico, Russia, Switzerland, and the United States of America.

5 See, e.g., Melissa Magliana, The Autonomous Province of South Tyrol: A Model for Self- Government (Bozen/Bolzano, Italy: European Academy of Bozen/Bolzano 2000); Francesco Palermo, “Self-Government (and other?) Instruments for the Prevention of Ethnic Conflicts in South Tyrol, ” in Mitia Zagar, Boris Jesih, and Romana Bester, eds., The Constitutional and Political Regulation of Ethnic Relations and Conflict (Institute for Ethnic Studies: Ljubljana, 1999); and Jens Woelk, “From Minority-Protection to Governance of Cohabitation? The Case of South Tyrol.” in Essays on Regionalism, ed. Agency of Local Democracy Subotica (Subotica: Yugoslavia, Agency of Local Democracy, 2001).

6 G. Alan Tarr, “Of Tribes and Territories: Asymmetry in American Federalism,” in Alain-G. Gagnon and Michael Burgess, eds., Revisiting Unity and Diversity in Federal Countries: Changing Concepts, Reform Proposals and New Institutional Realities (Leiden/Boston: Brill Njihoff, 2018).

7 In 1982, Daniel Elazar, observed that: “Students of federal systems have tended to focus their attention on the federal constitutions that frame the entire polity while neglecting the constitutional arrangements of the constituent polities. … In fact, the constitutions of constituent states are part and parcel of the total constitutional structure of federal systems and play a vital role in giving the system direction.” See Daniel J. Elazar, “The Principles and Traditions Underlying State Constitutions,” Publius: The Journal of Federalism 11 (Winter 1982): 18-22.

8 In this respect, the study of the sub-national perspective serves the same function as does the study of comparative constitutionalism more generally, namely, to broaden our understanding by making problematic and contingent what seemed obvious. As Kim Scheppele has put it: “One reason [for studying comparative constitutionalism] is that many of the taken-for-granted fixed starting points of our field are

23

actually variables connected to time and space, variables whose variable quality is obscured if we do not know the counterexamples.” Kim Lane Scheppele, “The Agendas of Comparative Constitutionalism,” Law and Courts (Spring 2003): 5.

9 This study focuses on the legal role of sub-national constitutions as independent sources of law. But this is not their only importance. Sub-national constitutions may serve important political purposes, regardless of the contents of the documents. They may be instruments of conflict management during periods of political stability, and the process of sub-national constitution-making itself may contribute to political socialization. Sub-national constitutions may also be important as vehicles for making political statements about the character of the federation. See, for example, the quasi-constitutional Bill 99 enacted by the National Assembly of Quebec in 2000. Sub-national constitutions may also be drafted in order to differentiate the constituent unit from other units within the federation—i.e., as a way of emphasizing asymmetry. This is hardly a comprehensive list of the political functions of sub-national constitutions and sub-national constitution-drafting, and that those who devise a sub-national constitution may be divided as to the purposes the constitution is designed to serve.

10 Watts, “Provinces, Lander, and Cantons,” at 945. On South Africa, see Bertus de Villiers, ed., Birth of a Constitution (Kenwyn: Juta & Co., 1994). On the Sudan, see Christina Murray and Catherine Maywald, “Subnational Constitution-Making in Southern Sudan,” Rutgers Law Journal 37 (Summer 2006): 1203- 1234. On Switzerland, see Giovanni Biaggini, “Federalism, Subnational Constitutional Arrangements, and the Protection of Minorities in Switzerland,” in Federalism, Subnational Constitutions, and Minority Rights.

11 Six states are exceptions to this statement. Texas was an independent republic before its annexation by the US, and five states--Vermont, Kentucky, Tennessee, Maine and West Virginia--were carved out of the territory of existing states. This represents a form of devolutionary federalism.

12 The main provision dealing with the admission of new states is Article IV, section 3 of the U.S. Constitution. Further constitutional support for congressional conditions on admission is provided by Article IV, sect. 4 of the U.S. Constitution, which directs the federal government to “guarantee to each State in the Union a Republican Form of Government.” In addition to imposing conditions on prospective states, Congress also supervised the constitutions that Southern states adopted in the aftermath of the Civil War, requiring an acceptable constitution as a condition for “readmission” to the Union.

For a discussion of somewhat similar processes in other countries, such as Switzerland and Spain, see Bertus de Villiers, "The Constitutional Principles: Content and Significance," in Bertus de Villiers, ed., Birth of a Constitution. (Kenwyn: Juta & Co., 1994).

14 For a discussion of this process of reunification and constitution-making, see Peter E. Quint, The Imperfect Union: Constitutional Structures of German Unification (Princeton, N.J.: Princeton University Press, 1997): 73-99.

15 Watts, “States, Provinces, Lander, and Cantons,” p. 945.

24

16 Anna Gamper writes: “Federal constitutions usually provide a large number of legal instruments in order to secure homogeneity among the various units. These include the determination of explicit constitutional rules that must neither be violated by the constitutions nor by the ordinary legislation or administration of the constituent states. The federal constitution may also determine that certain policies or law-making of the constituent units need the consent of the federation or an agreement between the units.” Gamper, “Austrian Federalism and the Protection of Minorities,” in Federalism, Subnational Constitutions, and Minority Rights, p. 59.

17 John Kincaid, “Comparative Observations,” and Juan Marcos Gutierrez Gonzalez, “United Mexican States,” in Constitutional Origins, Structure, and Change in Federal Countries, pp. 438 and 223. Under nineteenth-century Mexican constitutions, states had more leeway in designing their political institutions.

18 Austrian Basic Law, Art. 28, para. 1, and German Basic Law, Art. 28. However, the Federal Constitutional Court of Germany has noted that the lander “are states vested with their own sovereign powers…. The basic law requires only a certain degree of identity between the federal constitution and state constitutions. To the extent that the basic law [does not provide otherwise], the states are free to construct their own constitutional orders.” Quoted in Norman Weiss, “The Protection of Minorities in a Federal State: The Case of Germany,” in Federalism, Subnational Constitutions, and Minority Rights, p. 76.

19 Austrian Basic Law, Art. 99.

20 See, e.g., Reitman v. Mulkey, 387 U.S. 369 (1967); Hunter v. Underwood, 471 U.S. 222 (1985); Romer v. Evans, 517 U.S. 620 (1996).

21 See In re: Certification Interim Constitution of the Province of KwaZulu-Natal, 1996 1996 (11) BCLR 1419 (CC), at para. 15. For discussion of this litigation, see Robert F. Williams, “Comparative Subnational Constitutional Law: South Africa's Provincial Constitutional Experiments,” South Texas Law Review 40 (Summer 1999): 648-659,

22 Constitution of the Russian Federation, Article 85, section 2. According to State Prosecutor Yuri Skuratov, even prior to the accession of Vladimir Putin, the Justice Ministry revoked nearly 2000 regional laws. Quoted in "Constitution Watch," Eastern European Constitutional Review 7 (Winter 1998): 32.

23 See Tanja A. Bürzel, “From Competitive Regionalism to Cooperative Federalism: The Europeanization of the Spanish State of the Autonomies,” Publius: The Journal of Federalism 30 (Spring 2000): 17; Tanja A. Bürzel, States and Regions in the European Union: Institutional Adaptation in Germany and Spain (Cambridge: Cambridge University Press, 2002): 103-147.

24 The “principle of consideration” within the Austrian federation seems to play a similar role. See Gamper, “Austrian Federalism and the Protection of Rights,” pp. 62-63.

25 See Ronald L. Watts, Comparing Federal Systems, 3rd ed. (Montreal: McGill-Queen`s University Press, 2008).

26 Ibid., p. 244.

25

27 Stephen Tierney, Constitutional Law and National Pluralism (Oxford: Oxford University Press, 2004), p. 188.

28 See Vieytez, “Federalism, Subnational Constitutional Arrangements, and the Protection of Minorities in Spain,” p. 149.

29 South Africa Constitution, Ch. 4, sec. 44(2); India Constitution, secs. 249, 250.

30 See Austria Basic Law, Arts. 102(1), 103(1).

31 Celina Souza, “Federal Republic of Brazil,” in Constitutional Origins, Structure, and Change in Federal Countries, p. 85.

32 See Peter Bussjaeger and Mirella Johler, “Federalism and Recent Political Dynamics in Austria,” Revista d’estudis autonomics i federals 28 (2018).

33 There is an alternative version of this as well. Sometimes constituent units use their subnational constitutions to preserve what has been jettisoned at the national level. In the United States, for example, controls over liquor were maintained even after the federal government rejected prohibition with the ratification of the Twenty-first Amendment. In addition, state courts have interpreted state constitutions to maintain substantive due process after its repudiation by the U.S. Supreme Court and has recognized rights claims, such as the requirement of public funding for abortions, after the Supreme Court rejected such claims as a matter of federal constitutional law.

34 Arthur Gunlicks, “Land Constitutions in Germany,” Publius: The Journal of Federalism 28 (fall 1998): 111–12.

35 Gunlicks, “Land Constitutions in Germany,” pp.111-112.

36 See G. Alan Tarr, ed., Constitutional Politics in the States: Contemporary Controversies and Historical Patterns (Westport CT: Greenwood Press, 1996).

37 See Elazar, Exploring Federalism, pp. 18-22, and more generally, Daniel Elazar, American Federalism: A View from the States, 3rd ed. (New York: Harper & Row, 1984).

38 For discussion of these reforms, see Spencer C. Olin, Jr., California's Prodigal Sons; Hiram Johnson and the Progressives, 1911-1917 (Berkeley: University of California Press, 1968). .

39 See Rassie Malherbe and Dirk Brand, “South Africa: Sub-National Constitutional Law,” in Andre Alen et al., eds., International Encyclopedia of Laws, Constitutional Law, Sub-National Constitutional Law (Leiden: Kluwer Law International, 2001).

40 For studies focusing on the United States, see Jack L. Walker, “The Diffusion of Innovations Among the American States,” American Political Science Review 63 (September 1969): 880-889; Robert L. Savage, “Diffusion Research Traditions and the Spread of Policy Innovation in a Federal System, Publius: The Journal of Federalism 15 (Fall 1985): 1-27; and Virginia Gray, “Competition, Emulation, and Policy

26

Innovation,” in Lawrence C. Dodd and Calvin Jillson, eds., Perspectives in American Politics (Washington, DC: CQ Press, 1994).

41 James Willard Hurst, The Growth of American Law: The Law Makers (Boston: Little, Brown, 1950): 224- 225.

42 David Schuman, “The Origin of State Constitutional Direct Democracy: William Simon U'Ren and the ‘Oregon System’” Temple Law Review 67 (Fall 1994): 947, 950.

43 Obergefell v. Hodges, 576 U.S. 644 (2015).

27